Architecture

A standard Pigsty deployment comes with an INFRA module that provides the following services:

These are essential for a production-grade PostgreSQL service, and will be installed by default.

Component	Port	Domain	Description
`nginx`	`80`	`h.pigsty`	Web Service Portal (YUM/APT Repo)
`alertmanager`	`9093`	`a.pigsty`	Alert Aggregation and delivery
`prometheus`	`9090`	`p.pigsty`	Monitoring Time Series Database
`grafana`	`3000`	`g.pigsty`	Visualization Platform
`lok`	`3100`	-	Logging Collection Server
`pushgateway`	`9091`	-	Collect One-Time Job Metrics
`blackbox_exporter`	`9115`	-	ICMP, TCP, HTTP Probing
`dnsmasq`	`53`	-	DNS Server, optional
`chronyd`	`123`	-	NTP Time Server, optional

HA PG can be deployed without INFRA

If you don't want these, the Minimal Install mode deploy HA Postgres without the Infra module.

Nginx: Acts as a web server for local repo and a reverse proxy for other web UI services
Grafana : Visualization platform for metrics, dashboards, and data analytics
Loki: Centralized log aggregation and querying via Grafana
Prometheus: Time-series monitoring database for metrics collection, storage, and alert evaluation
AlertManager: Alert aggregation, notification dispatch, and silencing
PushGateway: Collects metrics from one-off and batch jobs
BlackboxExporter: Probes node IP and VIP reachability
DNSMASQ: DNS resolution for Pigsty’s internal domains
Chronyd: NTP time synchronization to keep all nodes in sync

The INFRA module isn’t mandatory for HA PostgreSQL - for instance, it’s omitted in Slim Install mode.

However, since INFRA provides essential supporting services for production-grade HA PostgreSQL clusters, it’s strongly recommended for most deployments.

If you already have your own infrastructure (Nginx, local repos, monitoring, DNS, NTP), you can disable INFRA and configure Pigsty to use your existing stack instead.

Nginx

Nginx is the gateway for all WebUI services in Pigsty, serving on HTTP (80) / HTTPS (443) by default.

It exposes web UIs like Grafana, Prometheus, AlertManager, and HAProxy console, while also serving static resources like local yum/apt repos.

Nginx configuration follows infra_portal definitions, for example:

infra_portal:
  home         : { domain: h.pigsty }
  grafana      : { domain: g.pigsty ,endpoint: "${admin_ip}:3000" ,websocket: true }
  prometheus   : { domain: p.pigsty ,endpoint: "${admin_ip}:9090" }
  alertmanager : { domain: a.pigsty ,endpoint: "${admin_ip}:9093" }
  blackbox     : { endpoint: "${admin_ip}:9115" }
  loki         : { endpoint: "${admin_ip}:3100" }
  #minio        : { domain: sss.pigsty  ,endpoint: "${admin_ip}:9001" ,scheme: https ,websocket: true }

These endpoint definitions are referenced by other services - logs go to loki endpoint, Grafana datasources register to grafana endpoint, alerts route to alertmanager endpoint.

Pigsty allows rich Nginx customization as a local file server or reverse proxy, with self-signed or real HTTPS certs.

For more details, check these tutorials:

Local Repo

During installation, Pigsty creates a local software repo on the INFRA node to speed up later software installations.

Located at /www/pigsty and served by Nginx, it’s accessible via http://h.pigsty/pigsty.

Pigsty’s offline package is a tarball of a pre-built repo directory. If /www/pigsty exists with a /www/pigsty/repo_complete marker, Pigsty skips downloading from upstream - perfect for air-gapped environments!

Repo definition lives in /www/pigsty.repo, fetchable via http://${admin_ip}/pigsty.repo:

curl -L http://h.pigsty/pigsty.repo -o /etc/yum.repos.d/pigsty.repo

You can also use the file repo directly without Nginx:

[pigsty-local]
name=Pigsty local $releasever - $basearch
baseurl=file:///www/pigsty/
enabled=1
gpgcheck=0

Local repo configs are in: Config: INFRA - REPO

Prometheus

Prometheus, our TSDB for monitoring, listens on port 9090 (access via IP:9090 or http://p.pigsty).

Key features:

Service discovery via local static files with identity info
Metric scraping, pre-processing, and TSDB storage
Alert rule evaluation and forwarding to AlertManager

AlertManager handles alerts on port 9093 (IP:9093 or http://a.pigsty). While it receives Prometheus alerts, you’ll need extra config (e.g., SMTP) for notifications.

Configs for Prometheus, AlertManager, PushGateway, and BlackboxExporter are in: Config: INFRA - PROMETHEUS

Grafana

Grafana, our visualization powerhouse, runs on port 3000 (IP:3000 or http://g.pigsty).

Pigsty’s monitoring is dashboard-based with URL-driven navigation. Drill down or up quickly to pinpoint issues.

Fun fact: We’ve supercharged Grafana with extra viz plugins like ECharts - it’s not just monitoring, it’s a low-code data app platform!

Loki handles logs on port 3100, with Promtail shipping logs from nodes to the mothership.

Configs live in: Config: INFRA - GRAFANA and Config: INFRA - Loki

Ansible

Ansible is already installed on the Admin Node during bootstrap. So the ansible on infra nodes are not actually used. But it gives you a viable backup option in case of your admin node is nuked.